[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-128432-en":3,"doc-seo-128432-105":31,"detail-sidebar-cat-0-en-105":92},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":21,"is_downloadable":21,"audit_status":21,"page_count":22,"language":23,"language_code":24,"site_id":25,"html_lang":24,"table_of_contents":26,"faqs":27,"seo_title":28,"seo_description":14,"update_tm":29,"read_time":30},128432,962084925290,"Ophelia","https://ap-avatar.wpscdn.com/davatar_085a072bc5b1113ac321206ff7593b45",8,"Research & Report","Machine learning for retrosynthesis and synthesisable molecule generation in drug discovery - Doctor of Philosophy thesis","Drug discovery is a slow, costly process with a declining success rate. Computer-Aided Drug Design can improve early-stage efficiency by evaluating more compounds per cycle and quickly pre-filtering molecules before synthesis, yet many in-silico proposals remain non-synthesisable or have unclear routes, wasting computational effort. This thesis develops methods to assess and improve synthesisability via retrosynthesis prediction and synthesisable, route-aware molecule generation.","Machine learning for retrosynthesis and synthesisable molecule generation in drug discovery  \nEwa Wieczorek  \nGreen Templeton College  \nUniversity of Oxford  \nA thesis submitted in partial fulfillment of the requirements for the  \ndegree of  \nDoctor of Philosophy  \nMichaelmas 2024  \nAcknowledgements  \nI would like to thank all my supervisors, both academic and industrial, for all the support they have given me. My journey in drug discovery has started with working in Paul’s lab, and it was great to have his input in my PhD research. Thank you to Fernanda for welcoming me to her group and letting me work on such interesting topics!  \nWorking closely with Exscientia has allowed me to see a different side ofcheminformatics research and put my work in perspective of the real-life drug discovery applications. I am very grateful to Liam for providing me with the opportunity to spend part of my PhD there and his mentorshiphis ideas and input into my research have been invaluable. Thank you also to Nina, Kirubin and Mason for welcoming me to the team and showing me the ropes!  \nI have enjoyed the three years I spent in the Duarte group immenselythank you to all the members for creating a great atmosphere. In particular my thanks go to the members of ML subgroup (past and present), including Chloe, Matina, Ally and Josh among others, for interesting discussions and input into my research. Also to all the members of the tiny office - Sara, Hanwen and Tom - thank you for all the chats and coffee breaks. You have made my last year at Oxford the most enjoyable!  \nFinally, thank you to all my friends and family who kept me sane and motivated - I would not have made it through the last 4 years without you!  \nAbstract  \nDrug discovery is a notoriously difficult and slow process, with high research and development costs and a decreasing success rate. ComputerAided Drug Design methods show promise in improving the efficiency of early stage drug discovery, increasing the number of compounds that can be evaluated per design cycle and allowing for pre-filtering of molecules with fast computational methods before they are synthesised. However, many of the compounds designed in silico are not synthesisable in practice or the synthesis routes towards them are not obvious. This leads to computational resources being wasted on designing molecules that can never be tested experimentally. This thesis explores new methods for two approaches assessing and improving synthesisability in drug discovery: retrosynthesis prediction and synthesisability-constrained molecule generation.  \nFirst, the problem of retrosynthesis prediction for molecules containing heterocyclic scaffolds is considered. Four domain adaptation approaches are benchmarked to develop a single-step retrosynthesis prediction model with improved performance for ring disconnections. Accuracy for heterocycle formations and all reaction classes, as well as computational cost, are considered. A further fine-tuning workflow for continual retraining of the model with newly published data is introduced. The application of the most versatile model, trained with a mixed fine-tuning strategy, is then demonstrated in multi-step retrosynthesis in a retrospective analysis for two drug-like compounds.  \nNext, the development of retro-active, a method for synthesisable molecule generation and optimisation, is described. Retro-active generates molecules based on a known synthesis route and a provided starting material pool. The use of active learning for starting material selection allows for the optimisation of the resulting product molecules for user-defined scoring  \nfunctions. A benchmark of starting material acquisition and product enumeration methods is included, as well as a comparison to alternative nonmachine learning-based starting material selection approaches. The applicability of retro-active for both ligand-based and structure-based drug discovery is demonstrated.  \nThe use case of retro-active ","cbCaimkaOk0uyo8l","https://ap.wps.com/l/cbCaimkaOk0uyo8l","pdf",23055059,2,1,181,"English","en",105,"# Introduction\n## Drug Discovery\n## Computer-Aided Drug Design\n## Synthetic accessibility prediction\n## Reaction prediction tasks\n## Computer-Aided Synthesis Planning\n## Thesis aims and outline\n# Theory\n## Chemical data representations\n## Machine learning","[{\"question\":\"Why is synthesisability a major challenge in drug discovery workflows?\",\"answer\":\"Many designed molecules cannot be synthesized in practice or lack clear synthetic routes, causing computational resources to be spent on compounds that cannot be experimentally tested.\"},{\"question\":\"What does the thesis cover first in improving synthesisability?\",\"answer\":\"It studies retrosynthesis prediction for molecules with heterocyclic scaffolds, benchmarking domain adaptation approaches, introducing continual fine-tuning, and applying the model in multi-step retrospective analysis for drug-like compounds.\"},{\"question\":\"How does Retro-active support synthesisable molecule generation?\",\"answer\":\"Retro-active generates molecules from a known synthesis route and a provided starting-material pool, using active learning for starting material selection and optimizing resulting products for user-defined scoring functions.\"}]","Machine learning for retrosynthesis and synthesisable molecule generation in drug discovery - Doctor of Philosophy thesis | PDF",1785947667,456,{"code":4,"msg":32,"data":33},"ok",{"site_id":25,"language":24,"slug":34,"title":13,"keywords":35,"description":14,"schema_data":36,"social_meta":87,"head_meta":89,"extra_data":91,"updated_unix":29},"machine-learning-for-retrosynthesis-and-synthesisable-molecule-generation-in-drug-discovery-doctor-of-philosophy-thesis","",{"@graph":37,"@context":86},[38,54,69],{"@type":39,"itemListElement":40},"BreadcrumbList",[41,45,48,51],{"item":42,"name":43,"@type":44,"position":21},"https://docshare.wps.com","Home","ListItem",{"item":46,"name":47,"@type":44,"position":20},"https://docshare.wps.com/document/","Document",{"item":49,"name":12,"@type":44,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":44,"position":53},"https://docshare.wps.com/document/machine-learning-for-retrosynthesis-and-synthesisable-molecule-generation-in-drug-discovery-doctor-of-philosophy-thesis/128432/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":24,"description":14,"dateModified":62,"datePublished":63,"encodingFormat":61,"isAccessibleForFree":64,"interactionStatistic":65},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":42,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-22","2026-08-05",true,{"@type":66,"interactionType":67,"userInteractionCount":20},"InteractionCounter",{"@type":68},"ViewAction",{"@type":70,"mainEntity":71},"FAQPage",[72,78,82],{"name":73,"@type":74,"acceptedAnswer":75},"Why is synthesisability a major challenge in drug discovery workflows?","Question",{"text":76,"@type":77},"Many designed molecules cannot be synthesized in practice or lack clear synthetic routes, causing computational resources to be spent on compounds that cannot be experimentally tested.","Answer",{"name":79,"@type":74,"acceptedAnswer":80},"What does the thesis cover first in improving synthesisability?",{"text":81,"@type":77},"It studies retrosynthesis prediction for molecules with heterocyclic scaffolds, benchmarking domain adaptation approaches, introducing continual fine-tuning, and applying the model in multi-step retrospective analysis for drug-like compounds.",{"name":83,"@type":74,"acceptedAnswer":84},"How does Retro-active support synthesisable molecule generation?",{"text":85,"@type":77},"Retro-active generates molecules from a known synthesis route and a provided starting-material pool, using active learning for starting material selection and optimizing resulting products for user-defined scoring functions.","https://schema.org",{"og:url":52,"og:type":88,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":90,"canonical":52},"index,follow",{"doc_id":7,"site_id":25},{"code":4,"msg":5,"data":93},[94,98,102,106,111,116,121,124,129,132,136],{"id":21,"doc_module":4,"doc_module_name":47,"category_name":95,"show_sort_weight":96,"slug":97},"Story & Novel",90,"story-novel",{"id":20,"doc_module":4,"doc_module_name":47,"category_name":99,"show_sort_weight":100,"slug":101},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":47,"category_name":103,"show_sort_weight":104,"slug":105},"Exam",70,"exam",{"id":107,"doc_module":4,"doc_module_name":47,"category_name":108,"show_sort_weight":109,"slug":110},5,"Comic",60,"comic",{"id":112,"doc_module":4,"doc_module_name":47,"category_name":113,"show_sort_weight":114,"slug":115},6,"Technology",50,"technology",{"id":117,"doc_module":4,"doc_module_name":47,"category_name":118,"show_sort_weight":119,"slug":120},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":47,"category_name":12,"show_sort_weight":122,"slug":123},30,"research-report",{"id":125,"doc_module":4,"doc_module_name":47,"category_name":126,"show_sort_weight":127,"slug":128},9,"Religion & Spirituality",20,"religion-spirituality",{"id":127,"doc_module":4,"doc_module_name":47,"category_name":130,"show_sort_weight":127,"slug":131},"World Cup","world-cup",{"id":133,"doc_module":4,"doc_module_name":47,"category_name":134,"show_sort_weight":133,"slug":135},10,"Lifestyle","lifestyle",{"id":137,"doc_module":4,"doc_module_name":47,"category_name":138,"show_sort_weight":107,"slug":139},19,"General","general"]